Shqipëria Posted on 2025-04-08 15:00:00

Artificial intelligence can prevent fires - New method identifies potential hot spots, analyzes risk factors

From Kristi Ceta

Artificial intelligence can prevent fires - New method identifies potential hot

A European climate research institute is using artificial intelligence (AI) to more accurately predict the probability of wildfires across the globe.

The European Centre for Medium-Range Weather Forecasts has created a new model that uses machine learning methods to analyze a wider set of data for more accurate predictions of where fires may break out.

Historically, fire forecasts have used a method called the Fire Weather Index, which uses four weather variables—temperature, wind, precipitation, and humidity. But there are limitations to this model.

"This doesn't account for a lot of detail," Climate Scientist Dr. Joe McNorton told Reuters. "It doesn't take into account flammable materials or ignition sources."

By including vegetation levels that can help spread flames and fire ignition sources in the analysis along with weather condition forecasts, the center can go further than just predicting fire risk.

"Machine learning really helps to get more accuracy in the location where a fire might occur," McNorton said.

As an example, scientist Francesca Di Giuseppe cited the devastating January fires in Los Angeles, which killed 28 people and damaged or destroyed nearly 16,000 structures.

Unusually wet weather conditions in the period before the fires caused abundant vegetation growth, which then served as fuel when the unusually dry winter weather began.

The Fire Probability Method identified the regions that could be most severely affected by the fires. “Our prediction in this case was much more accurate and pinpointed the exact location, very close to Los Angeles, where the fire actually occurred,” Giuseppe said.

Last year, record levels of greenhouse gases helped push temperatures to all-time highs and extreme weather caused widespread problems across the globe. Wildfires forced the displacement of 800,000 people, the highest number since records began in 2008.

McNorton said the AI-powered method could be used by fire prevention agencies to more accurately identify fire risk locations, allowing them to allocate resources appropriately. “It’s a big step forward in terms of fire prediction,” he said.

Live TV

Latest news
All news

Most visited